Vapor intrusion (VI) pathway assessment often involves the collection and analysis of groundwater, soil gas, and indoor air data. There is temporal variability in these data, but little is understood about the characteristics of that variability and how it influences pathway assessment decision-making. This research included the first-ever collection of a long-term high-frequency indoor air data set at a house with VI impacts overlying a dilute chlorinated solvent groundwater plume. It also included periodic synoptic snapshots of groundwater and soil gas data and high-frequency monitoring of building conditions and environmental factors. Indoor air trichloroethylene (TCE) concentrations varied over three orders-of-magnitude under natural conditions, with the highest daily VI activity during fall, winter, and spring months. These data were used to simulate outcomes from common sampling strategies, with the result being that there was a high probability (up to 100%) of false-negative decisions and poor characterization of long-term exposure. Temporal and spatial variability in subsurface data were shown to increase as the sampling point moves from source depth to ground surface, with variability of an order-of-magnitude or more for sub-slab soil gas. It was observed that indoor vapor sources can cause subsurface vapor clouds and that it can take days to weeks for soil gas plumes created by indoor sources to dissipate following indoor source removal. A long-term controlled pressure method (CPM) test was conducted to assess its utility as an alternate approach for VI pathway assessment. Indoor air concentrations were similar to maximum concentrations under natural conditions (9.3 μg/m3 average vs. 13 μg/m3 for 24 h TCE data) with little temporal variability. A key outcome was that there were no occurrences of false-negative results. Results suggest that CPM tests can produce worst-case exposure conditions at any time of the year. The results of these studies highlight the limitations of current VI pathway assessment approaches and demonstrate the need for robust alternate diagnostic tools, such as CPM, that lead to greater confidence in data interpretation and decision-making.